{smcl}
{com}{sf}{ul off}{txt}{.-}
      name:  {res}<unnamed>
       {txt}log:  {res}/Users/AliceZhang/Dropbox/Research_Columbia/Renewables Voting (Urpelainen Zhang)/JOP/UZ_JOP2021_Replication/Analysis/logSTATA/012_analysis_Texas.smcl
  {txt}log type:  {res}smcl
 {txt}opened on:  {res} 6 Nov 2021, 20:34:30
{txt}
{com}. 
. 
. *******************************************************************************
. /*                                        URPELAINEN & ZHANG 2021                                                */
. *******************************************************************************
. 
. /* 
> 
> File Name:      012_analysis_Texas.do
> 
> By:                             Alice Tianbo Zhang (alice.tianbo.zhang@gmail.com)
> 
> Last Edited:    10/11/2021
> 
> Purpose:                
> 
> Data Used:      votes_wind_panel_Texas.dta
> 
> Program Used:   - reghdfe -
> 
> */
. 
. 
. *******************************************************************************
. /*                                      TABLE A11                                                                    */
. *******************************************************************************
. 
. ** Load election district panel
. cd "$rootDir/$dataDir/Final"
{res}/Users/AliceZhang/Dropbox/Research_Columbia/Renewables Voting (Urpelainen Zhang)/JOP/UZ_JOP2021_Replication/Data/Final
{txt}
{com}. use election_district_panel_Texas.dta, clear
{txt}
{com}. 
. // Note: Election data from 2004 - 2010; corresponds to stable congressional district boundaries 
. 
. ** Create instrument and fixed effects
. gen t = year - 2004
{txt}
{com}. gen inter = t * mean_wp
{txt}
{com}. 
. egen stateyear_fixed = group(state year)
{txt}
{com}. egen district_fixed = group(state district)
{txt}
{com}. 
. gen cum_lncapacity_turbine = log(cum_capacity_turbine + 1)
{txt}
{com}. gen cum_lncount_turbine = log(cum_count_turbine +1 )
{txt}
{com}. 
. gen turnout = votes1 + votes2 + votes3 + votes4
{txt}
{com}. gen lnturnout = log(turnout)
{txt}(1 missing value generated)

{com}. 
. local outcome demvotesmajorpercent repvotesmajorpercent incumbvotesmajorpercent turnout lnturnout // outcome variable 
{txt}
{com}. local endogenous cum_capacity_turbine cum_count_turbine cum_lncapacity_turbine cum_lncount_turbine // endogenous variable
{txt}
{com}. local instrument inter // instrument
{txt}
{com}. local admin1_trend stateyear_fixed // geography * time trend
{txt}
{com}. local admin2 district_fixed // panel unit (cluster variable)
{txt}
{com}. 
. foreach y in `outcome' {c -(}
{txt}  2{com}.         // Create outcome variable label for storing estimates
.         local y_name = substr("`y'", 1, 3)
{txt}  3{com}.         
.         foreach x in `endogenous' {c -(}
{txt}  4{com}.                         // Create endogenous variable label for storing estimates
.                         tokenize "`x'", parse("_")
{txt}  5{com}.                         local x_name "`3'"
{txt}  6{com}.                         di "`x_name'"
{txt}  7{com}.                         
.                         // Run IV regression
.                         reghdfe `y' (`x' = `instrument'), absorb(`admin1_trend' `admin2') ffirst stages(first ols reduced) vce(cluster `admin2') old
{txt}  8{com}.                         
.                         // Store IV, first stage, OLS, reduced form estimates
.                         estimates store `y_name'_`x_name'_iv
{txt}  9{com}.                         estimates restore reghdfe_first1
{txt} 10{com}.                         estimates store `y_name'_`x_name'_first
{txt} 11{com}.                         estimates restore reghdfe_ols
{txt} 12{com}.                         estimates store `y_name'_`x_name'_ols
{txt} 13{com}.                         estimates restore reghdfe_reduced
{txt} 14{com}.                         estimates store `y_name'_`x_name'_reduced
{txt} 15{com}.                 {c )-}
{txt} 16{com}. {c )-}
capacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_capacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      7.67
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0060
{txt}{col 51}R-squared{col 67}= {res}    0.7341
{txt}{col 51}Adj R-squared{col 67}= {res}    0.6018
{txt}{col 51}Within R-sq.{col 67}= {res}    0.1020
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  180.8348

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_capaci~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 55.07583{col 26}{space 2} 19.88988{col 37}{space 1}    2.77{col 46}{space 3}0.006{col 54}{space 4} 15.94345{col 67}{space 3} 94.20821
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.72
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.3957
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8260
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0008
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5934

{txt}{ralign 86:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}demvotesmajorpercent{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2}  .001403{col 34}{space 2} .0016498{col 45}{space 1}    0.85{col 54}{space 3}0.396{col 62}{space 4}-.0018428{col 75}{space 3} .0046489
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}demvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .8652173{col 26}{space 2} .2734199{col 37}{space 1}    3.16{col 46}{space 3}0.002{col 54}{space 4} .3272767{col 67}{space 3} 1.403158
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_capacity{col 14}{txt}|{col 18}{res}    7.67{col 28}  0.0060{col 37}{txt}|{col 42}{res}    8.37{col 51}  0.0038{col 60}{txt}|{col 65}{res}    7.67

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}7.08   {col 61}{txt}P-val={res}0.0078

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  132.48
Kleibergen-Paap Wald rk F statistic{col 65}    7.67

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    4.17
{txt}{col 55}Prob > F      = {res}  0.0420
{txt}Total (centered) SS     = {res}  78105.0974{txt}{col 55}Centered R2   = {res}  0.8744
{txt}Total (uncentered) SS   = {res}  78105.0974{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 84364.48599{txt}{col 55}Root MSE      = {res}   9.974

{txt}{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}demvotesmajorpercent{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2} .0157096{col 34}{space 2}  .007695{col 45}{space 1}    2.04{col 54}{space 3}0.042{col 62}{space 4} .0005699{col 75}{space 3} .0308492
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   7.084
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0078
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 132.484
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   7.668
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_capacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
count
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_count_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      6.92
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0089
{txt}{col 51}R-squared{col 67}= {res}    0.9493
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9241
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0960
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  112.5290

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_count_~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 33.14829{col 26}{space 2} 12.59721{col 37}{space 1}    2.63{col 46}{space 3}0.009{col 54}{space 4} 8.363888{col 67}{space 3} 57.93268
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.66
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4184
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8260
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0007
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5938

{txt}{ralign 83:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}demvotesmajorpe~t{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2} .0021405{col 31}{space 2} .0026417{col 42}{space 1}    0.81{col 51}{space 3}0.418{col 59}{space 4}-.0030569{col 72}{space 3} .0073378
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}demvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .8652173{col 26}{space 2} .2734199{col 37}{space 1}    3.16{col 46}{space 3}0.002{col 54}{space 4} .3272767{col 67}{space 3} 1.403158
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_count_tu{col 14}{txt}|{col 18}{res}    6.92{col 28}  0.0089{col 37}{txt}|{col 42}{res}    7.56{col 51}  0.0060{col 60}{txt}|{col 65}{res}    6.92

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}6.49   {col 61}{txt}P-val={res}0.0108

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  123.94
Kleibergen-Paap Wald rk F statistic{col 65}    6.92

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    3.90
{txt}{col 55}Prob > F      = {res}  0.0490
{txt}Total (centered) SS     = {res}  78105.0974{txt}{col 55}Centered R2   = {res}  0.8737
{txt}Total (uncentered) SS   = {res}  78105.0974{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 84870.38651{txt}{col 55}Root MSE      = {res}      10

{txt}{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}demvotesmajorpe~t{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2} .0261014{col 31}{space 2} .0132101{col 42}{space 1}    1.98{col 51}{space 3}0.049{col 59}{space 4} .0001111{col 72}{space 3} .0520917
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   6.493
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0108
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 123.936
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   6.924
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_count_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncapacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncapacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     30.53
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.8798
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8200
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0554
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.7385

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncapa~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1617487{col 26}{space 2} .0292736{col 37}{space 1}    5.53{col 46}{space 3}0.000{col 54}{space 4} .1041543{col 67}{space 3} .2193432
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.02
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.8829
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8259
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5970

{txt}{ralign 88:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}  demvotesmajorpercent{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2}-.0614346{col 36}{space 2}  .416751{col 47}{space 1}   -0.15{col 56}{space 3}0.883{col 64}{space 4}-.8813722{col 77}{space 3}  .758503
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}demvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .8652173{col 26}{space 2} .2734199{col 37}{space 1}    3.16{col 46}{space 3}0.002{col 54}{space 4} .3272767{col 67}{space 3} 1.403158
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncapaci{col 14}{txt}|{col 18}{res}   30.53{col 28}  0.0000{col 37}{txt}|{col 42}{res}   33.33{col 51}  0.0000{col 60}{txt}|{col 65}{res}   30.53

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}22.67  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   68.51
Kleibergen-Paap Wald rk F statistic{col 65}   30.53

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    7.45
{txt}{col 55}Prob > F      = {res}  0.0067
{txt}Total (centered) SS     = {res}  78105.0974{txt}{col 55}Centered R2   = {res}  0.8624
{txt}Total (uncentered) SS   = {res}  78105.0974{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 92438.85623{txt}{col 55}Root MSE      = {res}   10.44

{txt}{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}  demvotesmajorpercent{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2} 5.349144{col 36}{space 2} 1.959913{col 47}{space 1}    2.73{col 56}{space 3}0.007{col 64}{space 4} 1.493109{col 77}{space 3}  9.20518
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  22.673
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  68.506
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  30.530
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncapacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncount
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncount_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     30.07
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.9137
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8708
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0566
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.6431

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncoun~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1423271{col 26}{space 2}  .025953{col 37}{space 1}    5.48{col 46}{space 3}0.000{col 54}{space 4} .0912658{col 67}{space 3} .1933885
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.00
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.9870
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8259
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5971

{txt}{ralign 85:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}demvotesmajorperc~t{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2}  .007496{col 33}{space 2} .4598424{col 44}{space 1}    0.02{col 53}{space 3}0.987{col 61}{space 4}-.8972218{col 74}{space 3} .9122138
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}demvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .8652173{col 26}{space 2} .2734199{col 37}{space 1}    3.16{col 46}{space 3}0.002{col 54}{space 4} .3272767{col 67}{space 3} 1.403158
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncount_{col 14}{txt}|{col 18}{res}   30.07{col 28}  0.0000{col 37}{txt}|{col 42}{res}   32.83{col 51}  0.0000{col 60}{txt}|{col 65}{res}   30.07

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}23.64  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   69.97
Kleibergen-Paap Wald rk F statistic{col 65}   30.07

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    7.40
{txt}{col 55}Prob > F      = {res}  0.0069
{txt}Total (centered) SS     = {res}  78105.0974{txt}{col 55}Centered R2   = {res}  0.8634
{txt}Total (uncentered) SS   = {res}  78105.0974{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res}  91807.0511{txt}{col 55}Root MSE      = {res}    10.4

{txt}{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}demvotesmajorperc~t{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2} 6.079075{col 33}{space 2} 2.235306{col 44}{space 1}    2.72{col 53}{space 3}0.007{col 61}{space 4} 1.681217{col 74}{space 3} 10.47693
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  23.639
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  69.965
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  30.075
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncount_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
capacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_capacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      7.67
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0060
{txt}{col 51}R-squared{col 67}= {res}    0.7341
{txt}{col 51}Adj R-squared{col 67}= {res}    0.6018
{txt}{col 51}Within R-sq.{col 67}= {res}    0.1020
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  180.8348

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_capaci~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 55.07583{col 26}{space 2} 19.88988{col 37}{space 1}    2.77{col 46}{space 3}0.006{col 54}{space 4} 15.94345{col 67}{space 3} 94.20821
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.72
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.3957
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8260
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0008
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5934

{txt}{ralign 86:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}repvotesmajorpercent{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2} -.001403{col 34}{space 2} .0016498{col 45}{space 1}   -0.85{col 54}{space 3}0.396{col 62}{space 4}-.0046489{col 75}{space 3} .0018428
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}repvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.8652173{col 26}{space 2} .2734199{col 37}{space 1}   -3.16{col 46}{space 3}0.002{col 54}{space 4}-1.403158{col 67}{space 3}-.3272767
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_capacity{col 14}{txt}|{col 18}{res}    7.67{col 28}  0.0060{col 37}{txt}|{col 42}{res}    8.37{col 51}  0.0038{col 60}{txt}|{col 65}{res}    7.67

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}7.08   {col 61}{txt}P-val={res}0.0078

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  132.48
Kleibergen-Paap Wald rk F statistic{col 65}    7.67

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    4.17
{txt}{col 55}Prob > F      = {res}  0.0420
{txt}Total (centered) SS     = {res} 78105.09811{txt}{col 55}Centered R2   = {res}  0.8744
{txt}Total (uncentered) SS   = {res} 78105.09811{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 84364.48612{txt}{col 55}Root MSE      = {res}   9.974

{txt}{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}repvotesmajorpercent{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2}-.0157096{col 34}{space 2}  .007695{col 45}{space 1}   -2.04{col 54}{space 3}0.042{col 62}{space 4}-.0308492{col 75}{space 3}-.0005699
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   7.084
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0078
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 132.484
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   7.668
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_capacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
count
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_count_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      6.92
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0089
{txt}{col 51}R-squared{col 67}= {res}    0.9493
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9241
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0960
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  112.5290

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_count_~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 33.14829{col 26}{space 2} 12.59721{col 37}{space 1}    2.63{col 46}{space 3}0.009{col 54}{space 4} 8.363888{col 67}{space 3} 57.93268
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.66
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4184
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8260
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0007
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5938

{txt}{ralign 83:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}repvotesmajorpe~t{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2}-.0021405{col 31}{space 2} .0026417{col 42}{space 1}   -0.81{col 51}{space 3}0.418{col 59}{space 4}-.0073378{col 72}{space 3} .0030569
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}repvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.8652173{col 26}{space 2} .2734199{col 37}{space 1}   -3.16{col 46}{space 3}0.002{col 54}{space 4}-1.403158{col 67}{space 3}-.3272767
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_count_tu{col 14}{txt}|{col 18}{res}    6.92{col 28}  0.0089{col 37}{txt}|{col 42}{res}    7.56{col 51}  0.0060{col 60}{txt}|{col 65}{res}    6.92

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}6.49   {col 61}{txt}P-val={res}0.0108

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  123.94
Kleibergen-Paap Wald rk F statistic{col 65}    6.92

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    3.90
{txt}{col 55}Prob > F      = {res}  0.0490
{txt}Total (centered) SS     = {res} 78105.09811{txt}{col 55}Centered R2   = {res}  0.8737
{txt}Total (uncentered) SS   = {res} 78105.09811{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res}  84870.3866{txt}{col 55}Root MSE      = {res}      10

{txt}{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}repvotesmajorpe~t{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2}-.0261014{col 31}{space 2} .0132101{col 42}{space 1}   -1.98{col 51}{space 3}0.049{col 59}{space 4}-.0520917{col 72}{space 3}-.0001111
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   6.493
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0108
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 123.936
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   6.924
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_count_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncapacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncapacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     30.53
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.8798
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8200
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0554
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.7385

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncapa~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1617487{col 26}{space 2} .0292736{col 37}{space 1}    5.53{col 46}{space 3}0.000{col 54}{space 4} .1041543{col 67}{space 3} .2193432
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.02
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.8829
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8259
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5970

{txt}{ralign 88:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}  repvotesmajorpercent{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2} .0614346{col 36}{space 2}  .416751{col 47}{space 1}    0.15{col 56}{space 3}0.883{col 64}{space 4}-.7585031{col 77}{space 3} .8813722
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}repvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.8652173{col 26}{space 2} .2734199{col 37}{space 1}   -3.16{col 46}{space 3}0.002{col 54}{space 4}-1.403158{col 67}{space 3}-.3272767
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncapaci{col 14}{txt}|{col 18}{res}   30.53{col 28}  0.0000{col 37}{txt}|{col 42}{res}   33.33{col 51}  0.0000{col 60}{txt}|{col 65}{res}   30.53

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}22.67  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   68.51
Kleibergen-Paap Wald rk F statistic{col 65}   30.53

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    7.45
{txt}{col 55}Prob > F      = {res}  0.0067
{txt}Total (centered) SS     = {res} 78105.09811{txt}{col 55}Centered R2   = {res}  0.8624
{txt}Total (uncentered) SS   = {res} 78105.09811{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 92438.85564{txt}{col 55}Root MSE      = {res}   10.44

{txt}{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}  repvotesmajorpercent{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2}-5.349144{col 36}{space 2} 1.959913{col 47}{space 1}   -2.73{col 56}{space 3}0.007{col 64}{space 4} -9.20518{col 77}{space 3}-1.493109
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  22.673
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  68.506
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  30.530
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncapacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncount
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncount_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     30.07
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.9137
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8708
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0566
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.6431

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncoun~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1423271{col 26}{space 2}  .025953{col 37}{space 1}    5.48{col 46}{space 3}0.000{col 54}{space 4} .0912658{col 67}{space 3} .1933885
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.00
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.9870
{txt}{col 51}R-squared{col 67}= {res}    0.8838
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8259
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5971

{txt}{ralign 85:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}repvotesmajorperc~t{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2} -.007496{col 33}{space 2} .4598424{col 44}{space 1}   -0.02{col 53}{space 3}0.987{col 61}{space 4}-.9122138{col 74}{space 3} .8972218
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     10.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0017
{txt}{col 51}R-squared{col 67}= {res}    0.8849
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8276
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0099
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    9.5493

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}repvotesma~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.8652173{col 26}{space 2} .2734199{col 37}{space 1}   -3.16{col 46}{space 3}0.002{col 54}{space 4}-1.403158{col 67}{space 3}-.3272767
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncount_{col 14}{txt}|{col 18}{res}   30.07{col 28}  0.0000{col 37}{txt}|{col 42}{res}   32.83{col 51}  0.0000{col 60}{txt}|{col 65}{res}   30.07

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}23.64  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   69.97
Kleibergen-Paap Wald rk F statistic{col 65}   30.07

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}  10.01{col 61}{txt}P-val={res}0.0017
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}  10.93{col 61}{txt}P-val={res}0.0009
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   9.15{col 61}{txt}P-val={res}0.0025

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    7.40
{txt}{col 55}Prob > F      = {res}  0.0069
{txt}Total (centered) SS     = {res} 78105.09811{txt}{col 55}Centered R2   = {res}  0.8634
{txt}Total (uncentered) SS   = {res} 78105.09811{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 91807.05052{txt}{col 55}Root MSE      = {res}    10.4

{txt}{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}repvotesmajorperc~t{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2}-6.079074{col 33}{space 2} 2.235306{col 44}{space 1}   -2.72{col 53}{space 3}0.007{col 61}{space 4}-10.47693{col 74}{space 3}-1.681217
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  23.639
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  69.965
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  30.075
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncount_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
capacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 8 singleton observations)
{res}{txt}(converged in 8 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_capacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      6.99
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0086
{txt}{col 51}R-squared{col 67}= {res}    0.7366
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5903
{txt}{col 51}Within R-sq.{col 67}= {res}    0.1023
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}  190.7800

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_capaci~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 57.62759{col 26}{space 2}  21.7983{col 37}{space 1}    2.64{col 46}{space 3}0.009{col 54}{space 4} 14.73945{col 67}{space 3} 100.5157
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      6.21
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0132
{txt}{col 51}R-squared{col 67}= {res}    0.6925
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5216
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0028
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9010

{txt}{ralign 86:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}incumbvotesmajorpe~t{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2}-.0026135{col 34}{space 2}  .001049{col 45}{space 1}   -2.49{col 54}{space 3}0.013{col 62}{space 4}-.0046773{col 75}{space 3}-.0005496
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}           97              97              0     {c |} 
      district_fixed {c |}            0             317            317 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      0.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.9423
{txt}{col 51}R-squared{col 67}= {res}    0.6916
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5202
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9149

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}incumbvote~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}  .025664{col 26}{space 2} .3541839{col 37}{space 1}    0.07{col 46}{space 3}0.942{col 54}{space 4}-.6711928{col 67}{space 3} .7225207
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   316{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   316{txt})
{res}cum_capacity{col 14}{txt}|{col 18}{res}    6.99{col 28}  0.0086{col 37}{txt}|{col 42}{res}    7.65{col 51}  0.0057{col 60}{txt}|{col 65}{res}    6.99

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}6.57   {col 61}{txt}P-val={res}0.0104

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  120.99
Kleibergen-Paap Wald rk F statistic{col 65}    6.99

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}316{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9423
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       317
{txt}Number of observations               N  = {res}      1160
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   317{txt}{col 55}Number of obs = {res}    1160
{txt}{col 55}F(  1,   316) = {res}    0.01
{txt}{col 55}Prob > F      = {res}  0.9427
{txt}Total (centered) SS     = {res} 73238.43614{txt}{col 55}Centered R2   = {res}  0.6913
{txt}Total (uncentered) SS   = {res} 73238.43614{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 73314.73681{txt}{col 55}Root MSE      = {res}    9.92

{txt}{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}incumbvotesmajorpe~t{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2} .0004453{col 34}{space 2} .0061879{col 45}{space 1}    0.07{col 54}{space 3}0.943{col 62}{space 4}-.0117293{col 75}{space 3}   .01262
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   6.571
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0104
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 120.994
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   6.989
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_capacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}           97              97              0     {c |} 
      district_fixed {c |}            0             317            317 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
count
{err}(running historical version of reghdfe)
{res}{txt}(dropped 8 singleton observations)
{res}{txt}(converged in 8 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_count_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      6.36
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0121
{txt}{col 51}R-squared{col 67}= {res}    0.9458
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9157
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0971
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}  118.8511

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_count_~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 34.88243{col 26}{space 2} 13.82682{col 37}{space 1}    2.52{col 46}{space 3}0.012{col 54}{space 4} 7.678171{col 67}{space 3} 62.08669
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      7.67
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0060
{txt}{col 51}R-squared{col 67}= {res}    0.6925
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5216
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0030
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9003

{txt}{ralign 83:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}incumbvotesmajo~t{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2}-.0043092{col 31}{space 2} .0015564{col 42}{space 1}   -2.77{col 51}{space 3}0.006{col 59}{space 4}-.0073714{col 72}{space 3} -.001247
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}           97              97              0     {c |} 
   district_fixed {c |}            0             317            317 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      0.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.9423
{txt}{col 51}R-squared{col 67}= {res}    0.6916
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5202
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9149

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}incumbvote~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}  .025664{col 26}{space 2} .3541839{col 37}{space 1}    0.07{col 46}{space 3}0.942{col 54}{space 4}-.6711928{col 67}{space 3} .7225207
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   316{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   316{txt})
{res}cum_count_tu{col 14}{txt}|{col 18}{res}    6.36{col 28}  0.0121{col 37}{txt}|{col 42}{res}    6.97{col 51}  0.0083{col 60}{txt}|{col 65}{res}    6.36

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}6.06   {col 61}{txt}P-val={res}0.0139

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  114.23
Kleibergen-Paap Wald rk F statistic{col 65}    6.36

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}316{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9423
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       317
{txt}Number of observations               N  = {res}      1160
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   317{txt}{col 55}Number of obs = {res}    1160
{txt}{col 55}F(  1,   316) = {res}    0.01
{txt}{col 55}Prob > F      = {res}  0.9427
{txt}Total (centered) SS     = {res} 73238.43614{txt}{col 55}Centered R2   = {res}  0.6913
{txt}Total (uncentered) SS   = {res} 73238.43614{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 73318.65005{txt}{col 55}Root MSE      = {res}    9.92

{txt}{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}incumbvotesmajo~t{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2} .0007357{col 31}{space 2}  .010225{col 42}{space 1}    0.07{col 51}{space 3}0.943{col 59}{space 4}-.0193819{col 72}{space 3} .0208533
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   6.057
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0139
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 114.228
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   6.365
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_count_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}           97              97              0     {c |} 
   district_fixed {c |}            0             317            317 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncapacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 8 singleton observations)
{res}{txt}(converged in 8 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncapacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}     25.45
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.8872
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8245
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0548
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    0.7303

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncapa~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}  .157393{col 26}{space 2} .0311975{col 37}{space 1}    5.05{col 46}{space 3}0.000{col 54}{space 4}  .096012{col 67}{space 3} .2187741
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      1.57
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.2110
{txt}{col 51}R-squared{col 67}= {res}    0.6921
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5210
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0017
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9064

{txt}{ralign 88:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}incumbvotesmajorperc~t{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2}-.5470853{col 36}{space 2} .4365131{col 47}{space 1}   -1.25{col 56}{space 3}0.211{col 64}{space 4}-1.405925{col 77}{space 3} .3117539
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}           97              97              0     {c |} 
        district_fixed {c |}            0             317            317 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      0.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.9423
{txt}{col 51}R-squared{col 67}= {res}    0.6916
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5202
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9149

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}incumbvote~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}  .025664{col 26}{space 2} .3541839{col 37}{space 1}    0.07{col 46}{space 3}0.942{col 54}{space 4}-.6711928{col 67}{space 3} .7225207
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   316{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   316{txt})
{res}cum_lncapaci{col 14}{txt}|{col 18}{res}   25.45{col 28}  0.0000{col 37}{txt}|{col 42}{res}   27.87{col 51}  0.0000{col 60}{txt}|{col 65}{res}   25.45

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}20.16  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   61.59
Kleibergen-Paap Wald rk F statistic{col 65}   25.45

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}316{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9423
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       317
{txt}Number of observations               N  = {res}      1160
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   317{txt}{col 55}Number of obs = {res}    1160
{txt}{col 55}F(  1,   316) = {res}    0.01
{txt}{col 55}Prob > F      = {res}  0.9424
{txt}Total (centered) SS     = {res} 73238.43614{txt}{col 55}Centered R2   = {res}  0.6912
{txt}Total (uncentered) SS   = {res} 73238.43614{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 73324.61516{txt}{col 55}Root MSE      = {res}   9.921

{txt}{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}incumbvotesmajorperc~t{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2} .1630565{col 36}{space 2} 2.254174{col 47}{space 1}    0.07{col 56}{space 3}0.942{col 64}{space 4}-4.272029{col 77}{space 3} 4.598142
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  20.157
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  61.593
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  25.453
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncapacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}           97              97              0     {c |} 
        district_fixed {c |}            0             317            317 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncount
{err}(running historical version of reghdfe)
{res}{txt}(dropped 8 singleton observations)
{res}{txt}(converged in 8 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncount_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}     25.46
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.9189
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8738
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0560
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    0.6370

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncoun~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1388857{col 26}{space 2} .0275229{col 37}{space 1}    5.05{col 46}{space 3}0.000{col 54}{space 4} .0847345{col 67}{space 3}  .193037
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      0.87
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.3530
{txt}{col 51}R-squared{col 67}= {res}    0.6919
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5207
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0009
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9104

{txt}{ralign 85:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}incumbvotesmajorp~t{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2}-.4613032{col 33}{space 2} .4959473{col 44}{space 1}   -0.93{col 53}{space 3}0.353{col 61}{space 4}-1.437079{col 74}{space 3} .5144728
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}           97              97              0     {c |} 
     district_fixed {c |}            0             317            317 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,160
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    316{txt}){col 67}= {res}      0.01
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.9423
{txt}{col 51}R-squared{col 67}= {res}    0.6916
{txt}{col 51}Adj R-squared{col 67}= {res}    0.5202
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0000
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       317{txt}{col 51}Root MSE{col 67}= {res}    9.9149

{txt}{ralign 78:(Std. Err. adjusted for {res:317} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}incumbvote~t{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}  .025664{col 26}{space 2} .3541839{col 37}{space 1}    0.07{col 46}{space 3}0.942{col 54}{space 4}-.6711928{col 67}{space 3} .7225207
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}           97              97              0     {c |} 
 district_fixed {c |}            0             317            317 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   316{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   316{txt})
{res}cum_lncount_{col 14}{txt}|{col 18}{res}   25.46{col 28}  0.0000{col 37}{txt}|{col 42}{res}   27.88{col 51}  0.0000{col 60}{txt}|{col 65}{res}   25.46

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}21.35  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   63.04
Kleibergen-Paap Wald rk F statistic{col 65}   25.46

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}316{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9423
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.01{col 61}{txt}P-val={res}0.9396

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       317
{txt}Number of observations               N  = {res}      1160
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   317{txt}{col 55}Number of obs = {res}    1160
{txt}{col 55}F(  1,   316) = {res}    0.01
{txt}{col 55}Prob > F      = {res}  0.9424
{txt}Total (centered) SS     = {res} 73238.43614{txt}{col 55}Centered R2   = {res}  0.6913
{txt}Total (uncentered) SS   = {res} 73238.43614{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 73303.96196{txt}{col 55}Root MSE      = {res}   9.919

{txt}{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}incumbvotesmajorp~t{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2} .1847847{col 33}{space 2} 2.554347{col 44}{space 1}    0.07{col 53}{space 3}0.942{col 61}{space 4}-4.840892{col 74}{space 3} 5.210462
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  21.351
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  63.043
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  25.464
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncount_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}           97              97              0     {c |} 
     district_fixed {c |}            0             317            317 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
capacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_capacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      7.67
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0060
{txt}{col 51}R-squared{col 67}= {res}    0.7341
{txt}{col 51}Adj R-squared{col 67}= {res}    0.6018
{txt}{col 51}Within R-sq.{col 67}= {res}    0.1020
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  180.8348

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_capaci~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 55.07583{col 26}{space 2} 19.88988{col 37}{space 1}    2.77{col 46}{space 3}0.006{col 54}{space 4} 15.94345{col 67}{space 3} 94.20821
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.23
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.2683
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9456
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0004
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16044.5392

{txt}{ralign 86:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}             turnout{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2}-1.601548{col 34}{space 2} 1.444328{col 45}{space 1}   -1.11{col 54}{space 3}0.268{col 62}{space 4}-4.443194{col 75}{space 3} 1.240097
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.98
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.1599
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9457
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0019
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16032.0739

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-634.8378{col 26}{space 2} 450.6616{col 37}{space 1}   -1.41{col 46}{space 3}0.160{col 54}{space 4}-1521.493{col 67}{space 3} 251.8173
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_capacity{col 14}{txt}|{col 18}{res}    7.67{col 28}  0.0060{col 37}{txt}|{col 42}{res}    8.37{col 51}  0.0038{col 60}{txt}|{col 65}{res}    7.67

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}7.08   {col 61}{txt}P-val={res}0.0078

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  132.48
Kleibergen-Paap Wald rk F statistic{col 65}    7.67

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   1.98{col 61}{txt}P-val={res}0.1599
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.17{col 61}{txt}P-val={res}0.1411
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.01{col 61}{txt}P-val={res}0.1564

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    1.48
{txt}{col 55}Prob > F      = {res}  0.2252
{txt}Total (centered) SS     = {res} 2.18378e+11{txt}{col 55}Centered R2   = {res}  0.9632
{txt}Total (uncentered) SS   = {res} 2.18378e+11{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 2.21340e+11{txt}{col 55}Root MSE      = {res}   16156

{txt}{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}             turnout{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2}-11.52661{col 34}{space 2} 9.485134{col 45}{space 1}   -1.22{col 54}{space 3}0.225{col 62}{space 4}-30.18816{col 75}{space 3} 7.134932
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   7.084
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0078
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 132.484
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   7.668
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_capacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
count
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_count_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      6.92
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0089
{txt}{col 51}R-squared{col 67}= {res}    0.9493
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9241
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0960
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  112.5290

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_count_~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 33.14829{col 26}{space 2} 12.59721{col 37}{space 1}    2.63{col 46}{space 3}0.009{col 54}{space 4} 8.363888{col 67}{space 3} 57.93268
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.47
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.2267
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9456
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0004
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16044.3558

{txt}{ralign 83:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}          turnout{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2}-2.662287{col 31}{space 2} 2.198033{col 42}{space 1}   -1.21{col 51}{space 3}0.227{col 59}{space 4}-6.986812{col 72}{space 3} 1.662238
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.98
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.1599
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9457
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0019
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16032.0739

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-634.8378{col 26}{space 2} 450.6616{col 37}{space 1}   -1.41{col 46}{space 3}0.160{col 54}{space 4}-1521.493{col 67}{space 3} 251.8173
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_count_tu{col 14}{txt}|{col 18}{res}    6.92{col 28}  0.0089{col 37}{txt}|{col 42}{res}    7.56{col 51}  0.0060{col 60}{txt}|{col 65}{res}    6.92

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}6.49   {col 61}{txt}P-val={res}0.0108

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  123.94
Kleibergen-Paap Wald rk F statistic{col 65}    6.92

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   1.98{col 61}{txt}P-val={res}0.1599
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.17{col 61}{txt}P-val={res}0.1411
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.01{col 61}{txt}P-val={res}0.1564

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    1.45
{txt}{col 55}Prob > F      = {res}  0.2291
{txt}Total (centered) SS     = {res} 2.18378e+11{txt}{col 55}Centered R2   = {res}  0.9631
{txt}Total (uncentered) SS   = {res} 2.18378e+11{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 2.21523e+11{txt}{col 55}Root MSE      = {res}   16163

{txt}{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}          turnout{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2}-19.15145{col 31}{space 2} 15.89194{col 42}{space 1}   -1.21{col 51}{space 3}0.229{col 59}{space 4}-50.41809{col 72}{space 3} 12.11518
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   6.493
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0108
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 123.936
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   6.924
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_count_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncapacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncapacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     30.53
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.8798
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8200
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0554
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.7385

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncapa~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1617487{col 26}{space 2} .0292736{col 37}{space 1}    5.53{col 46}{space 3}0.000{col 54}{space 4} .1041543{col 67}{space 3} .2193432
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.11
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.2918
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9456
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0010
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16039.4978

{txt}{ralign 88:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}               turnout{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2}-664.7056{col 36}{space 2} 629.5303{col 47}{space 1}   -1.06{col 56}{space 3}0.292{col 64}{space 4}-1903.276{col 77}{space 3} 573.8651
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.98
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.1599
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9457
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0019
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16032.0739

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-634.8378{col 26}{space 2} 450.6616{col 37}{space 1}   -1.41{col 46}{space 3}0.160{col 54}{space 4}-1521.493{col 67}{space 3} 251.8173
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncapaci{col 14}{txt}|{col 18}{res}   30.53{col 28}  0.0000{col 37}{txt}|{col 42}{res}   33.33{col 51}  0.0000{col 60}{txt}|{col 65}{res}   30.53

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}22.67  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   68.51
Kleibergen-Paap Wald rk F statistic{col 65}   30.53

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   1.98{col 61}{txt}P-val={res}0.1599
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.17{col 61}{txt}P-val={res}0.1411
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.01{col 61}{txt}P-val={res}0.1564

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    1.88
{txt}{col 55}Prob > F      = {res}  0.1709
{txt}Total (centered) SS     = {res} 2.18378e+11{txt}{col 55}Centered R2   = {res}  0.9628
{txt}Total (uncentered) SS   = {res} 2.18378e+11{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 2.23366e+11{txt}{col 55}Root MSE      = {res}   16230

{txt}{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}               turnout{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2} -3924.84{col 36}{space 2} 2859.521{col 47}{space 1}   -1.37{col 56}{space 3}0.171{col 64}{space 4} -9550.81{col 77}{space 3} 1701.131
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  22.673
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  68.506
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  30.530
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncapacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncount
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 5 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncount_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     30.07
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.9137
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8708
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0566
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.6431

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncoun~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1423271{col 26}{space 2}  .025953{col 37}{space 1}    5.48{col 46}{space 3}0.000{col 54}{space 4} .0912658{col 67}{space 3} .1933885
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.42
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.2341
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9456
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0013
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16037.2143

{txt}{ralign 85:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}            turnout{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2}-865.5782{col 33}{space 2} 726.0262{col 44}{space 1}   -1.19{col 53}{space 3}0.234{col 61}{space 4}    -2294{col 74}{space 3} 562.8434
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,271
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      1.98
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.1599
{txt}{col 51}R-squared{col 67}= {res}    0.9637
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9457
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0019
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}16032.0739

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}     turnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-634.8378{col 26}{space 2} 450.6616{col 37}{space 1}   -1.41{col 46}{space 3}0.160{col 54}{space 4}-1521.493{col 67}{space 3} 251.8173
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncount_{col 14}{txt}|{col 18}{res}   30.07{col 28}  0.0000{col 37}{txt}|{col 42}{res}   32.83{col 51}  0.0000{col 60}{txt}|{col 65}{res}   30.07

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}23.64  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   69.97
Kleibergen-Paap Wald rk F statistic{col 65}   30.07

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   1.98{col 61}{txt}P-val={res}0.1599
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.17{col 61}{txt}P-val={res}0.1411
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   2.01{col 61}{txt}P-val={res}0.1564

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1271
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1271
{txt}{col 55}F(  1,   318) = {res}    1.88
{txt}{col 55}Prob > F      = {res}  0.1710
{txt}Total (centered) SS     = {res} 2.18378e+11{txt}{col 55}Centered R2   = {res}  0.9629
{txt}Total (uncentered) SS   = {res} 2.18378e+11{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 2.22902e+11{txt}{col 55}Root MSE      = {res}   16213

{txt}{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}            turnout{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2}-4460.413{col 33}{space 2} 3251.065{col 44}{space 1}   -1.37{col 53}{space 3}0.171{col 61}{space 4}-10856.73{col 74}{space 3} 1935.901
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  23.639
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  69.965
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  30.075
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncount_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
capacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 6 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_capacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      7.67
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0060
{txt}{col 51}R-squared{col 67}= {res}    0.7340
{txt}{col 51}Adj R-squared{col 67}= {res}    0.6015
{txt}{col 51}Within R-sq.{col 67}= {res}    0.1020
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  180.9400

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_capaci~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 55.08724{col 26}{space 2} 19.89712{col 37}{space 1}    2.77{col 46}{space 3}0.006{col 54}{space 4} 15.94062{col 67}{space 3} 94.23386
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      2.03
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.1551
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0005
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 86:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}           lnturnout{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2}-9.51e-06{col 34}{space 2} 6.67e-06{col 45}{space 1}   -1.43{col 54}{space 3}0.155{col 62}{space 4}-.0000226{col 75}{space 3} 3.62e-06
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.63
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4294
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0004
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   lnturnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.0015197{col 26}{space 2} .0019208{col 37}{space 1}   -0.79{col 46}{space 3}0.429{col 54}{space 4}-.0052987{col 67}{space 3} .0022594
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_capacity{col 14}{txt}|{col 18}{res}    7.67{col 28}  0.0060{col 37}{txt}|{col 42}{res}    8.37{col 51}  0.0038{col 60}{txt}|{col 65}{res}    7.67

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}7.08   {col 61}{txt}P-val={res}0.0078

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  132.39
Kleibergen-Paap Wald rk F statistic{col 65}    7.67

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   0.63{col 61}{txt}P-val={res}0.4294
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.68{col 61}{txt}P-val={res}0.4084
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.62{col 61}{txt}P-val={res}0.4296

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1270
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1270
{txt}{col 55}F(  1,   318) = {res}    0.56
{txt}{col 55}Prob > F      = {res}  0.4552
{txt}Total (centered) SS     = {res} 5.371855681{txt}{col 55}Centered R2   = {res}  0.9641
{txt}Total (uncentered) SS   = {res} 5.371855681{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 5.379159705{txt}{col 55}Root MSE      = {res}  .07969

{txt}{hline 21}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 22}{c |}{col 34}    Robust
{col 1}           lnturnout{col 22}{c |}      Coef.{col 34}   Std. Err.{col 46}      t{col 54}   P>|t|{col 62}     [95% Con{col 75}f. Interval]
{hline 21}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_capacity_turbine {c |}{col 22}{res}{space 2}-.0000276{col 34}{space 2} .0000369{col 45}{space 1}   -0.75{col 54}{space 3}0.455{col 62}{space 4}-.0001002{col 75}{space 3}  .000045
{txt}{hline 21}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   7.077
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0078
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 132.390
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   7.665
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_capacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 21}{c TT}{hline 49}{c TRC}
         Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 21}{c +}{hline 49}{c RT}
     stateyear_fixed {c |}          103             103              0     {c |} 
      district_fixed {c |}            0             319            319 *   {c |} 
{hline 21}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
count
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 6 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_count_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      6.92
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0090
{txt}{col 51}R-squared{col 67}= {res}    0.9493
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9240
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0960
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}  112.5953

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_count_~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} 33.14609{col 26}{space 2} 12.60246{col 37}{space 1}    2.63{col 46}{space 3}0.009{col 54}{space 4} 8.351347{col 67}{space 3} 57.94082
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      2.10
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.1484
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0005
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 83:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}        lnturnout{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2}-.0000144{col 31}{space 2} 9.94e-06{col 42}{space 1}   -1.45{col 51}{space 3}0.148{col 59}{space 4} -.000034{col 72}{space 3} 5.16e-06
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.63
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4294
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0004
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   lnturnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.0015197{col 26}{space 2} .0019208{col 37}{space 1}   -0.79{col 46}{space 3}0.429{col 54}{space 4}-.0052987{col 67}{space 3} .0022594
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_count_tu{col 14}{txt}|{col 18}{res}    6.92{col 28}  0.0090{col 37}{txt}|{col 42}{res}    7.55{col 51}  0.0060{col 60}{txt}|{col 65}{res}    6.92

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}6.48   {col 61}{txt}P-val={res}0.0109

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}  123.78
Kleibergen-Paap Wald rk F statistic{col 65}    6.92

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   0.63{col 61}{txt}P-val={res}0.4294
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.68{col 61}{txt}P-val={res}0.4084
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.62{col 61}{txt}P-val={res}0.4296

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1270
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1270
{txt}{col 55}F(  1,   318) = {res}    0.55
{txt}{col 55}Prob > F      = {res}  0.4577
{txt}Total (centered) SS     = {res} 5.371855681{txt}{col 55}Centered R2   = {res}  0.9641
{txt}Total (uncentered) SS   = {res} 5.371855681{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 5.381133006{txt}{col 55}Root MSE      = {res}  .07971

{txt}{hline 18}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 19}{c |}{col 31}    Robust
{col 1}        lnturnout{col 19}{c |}      Coef.{col 31}   Std. Err.{col 43}      t{col 51}   P>|t|{col 59}     [95% Con{col 72}f. Interval]
{hline 18}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_count_turbine {c |}{col 19}{res}{space 2}-.0000458{col 31}{space 2} .0000617{col 42}{space 1}   -0.74{col 51}{space 3}0.458{col 59}{space 4}-.0001672{col 72}{space 3} .0000755
{txt}{hline 18}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}   6.483
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0109
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71} 123.778
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}   6.918
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_count_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 18}{c TT}{hline 49}{c TRC}
      Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 18}{c +}{hline 49}{c RT}
  stateyear_fixed {c |}          103             103              0     {c |} 
   district_fixed {c |}            0             319            319 *   {c |} 
{hline 18}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncapacity
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 6 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncapacity_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     30.25
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.8793
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8192
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0552
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.7383

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncapa~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1612546{col 26}{space 2} .0293183{col 37}{space 1}    5.50{col 46}{space 3}0.000{col 54}{space 4} .1035723{col 67}{space 3}  .218937
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.50
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4809
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0003
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 88:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}             lnturnout{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2}-.0018868{col 36}{space 2} .0026735{col 47}{space 1}   -0.71{col 56}{space 3}0.481{col 64}{space 4}-.0071468{col 77}{space 3} .0033732
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.63
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4294
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0004
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   lnturnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.0015197{col 26}{space 2} .0019208{col 37}{space 1}   -0.79{col 46}{space 3}0.429{col 54}{space 4}-.0052987{col 67}{space 3} .0022594
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncapaci{col 14}{txt}|{col 18}{res}   30.25{col 28}  0.0000{col 37}{txt}|{col 42}{res}   33.03{col 51}  0.0000{col 60}{txt}|{col 65}{res}   30.25

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}22.34  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   68.14
Kleibergen-Paap Wald rk F statistic{col 65}   30.25

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   0.63{col 61}{txt}P-val={res}0.4294
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.68{col 61}{txt}P-val={res}0.4084
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.62{col 61}{txt}P-val={res}0.4296

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1270
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1270
{txt}{col 55}F(  1,   318) = {res}    0.63
{txt}{col 55}Prob > F      = {res}  0.4272
{txt}Total (centered) SS     = {res} 5.371855681{txt}{col 55}Centered R2   = {res}  0.9640
{txt}Total (uncentered) SS   = {res} 5.371855681{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res}  5.39787538{txt}{col 55}Root MSE      = {res}  .07983

{txt}{hline 23}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 24}{c |}{col 36}    Robust
{col 1}             lnturnout{col 24}{c |}      Coef.{col 36}   Std. Err.{col 48}      t{col 56}   P>|t|{col 64}     [95% Con{col 77}f. Interval]
{hline 23}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncapacity_turbine {c |}{col 24}{res}{space 2} -.009424{col 36}{space 2} .0118544{col 47}{space 1}   -0.79{col 56}{space 3}0.427{col 64}{space 4} -.032747{col 77}{space 3} .0138991
{txt}{hline 23}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  22.344
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  68.140
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  30.251
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncapacity_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 23}{c TT}{hline 49}{c TRC}
           Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 23}{c +}{hline 49}{c RT}
       stateyear_fixed {c |}          103             103              0     {c |} 
        district_fixed {c |}            0             319            319 *   {c |} 
{hline 23}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)
lncount
{err}(running historical version of reghdfe)
{res}{txt}(dropped 1 singleton observations)
{res}{txt}(converged in 6 iterations)
{res}
{txt}{inp}{title:Stage: first - cum_lncount_turbine}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}     29.74
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.0000
{txt}{col 51}R-squared{col 67}= {res}    0.9136
{txt}{col 51}Adj R-squared{col 67}= {res}    0.8705
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0563
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.6425

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}cum_lncoun~e{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2} .1417991{col 26}{space 2} .0260008{col 37}{space 1}    5.45{col 46}{space 3}0.000{col 54}{space 4} .0906438{col 67}{space 3} .1929544
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: ols}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.59
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4442
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0003
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 85:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}          lnturnout{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2} -.002237{col 33}{space 2} .0029203{col 44}{space 1}   -0.77{col 53}{space 3}0.444{col 61}{space 4}-.0079825{col 74}{space 3} .0035085
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: reduced}

{txt}HDFE Linear regression{col 51}Number of obs{col 67}= {res}     1,270
{txt}Absorbing 2 HDFE groups{col 51}F({res}   1{txt},{res}    318{txt}){col 67}= {res}      0.63
{txt}Statistics robust to heteroskedasticity{col 51}Prob > F{col 67}= {res}    0.4294
{txt}{col 51}R-squared{col 67}= {res}    0.9642
{txt}{col 51}Adj R-squared{col 67}= {res}    0.9463
{txt}{col 51}Within R-sq.{col 67}= {res}    0.0004
{txt}{col 1}Number of clusters ({res}district_fixed{txt}) {col 30}= {res}       319{txt}{col 51}Root MSE{col 67}= {res}    0.0796

{txt}{ralign 78:(Std. Err. adjusted for {res:319} clusters in district_fixed)}
{hline 13}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 14}{c |}{col 26}    Robust
{col 1}   lnturnout{col 14}{c |}      Coef.{col 26}   Std. Err.{col 38}      t{col 46}   P>|t|{col 54}     [95% Con{col 67}f. Interval]
{hline 13}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 7}inter {c |}{col 14}{res}{space 2}-.0015197{col 26}{space 2} .0019208{col 37}{space 1}   -0.79{col 46}{space 3}0.429{col 54}{space 4}-.0052987{col 67}{space 3} .0022594
{txt}{hline 13}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}

Absorbed degrees of freedom:
{hline 16}{c TT}{hline 49}{c TRC}
    Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 16}{c +}{hline 49}{c RT}
stateyear_fixed {c |}          103             103              0     {c |} 
 district_fixed {c |}            0             319            319 *   {c |} 
{hline 16}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}
{txt}{inp}{title:Stage: iv}


{txt}Summary results for first-stage regressions
{hline 43}

{col 44}{help ivreg2##swstats:(Underid)}{col 65}{help ivreg2##swstats:(Weak id)}
Variable     |{col 16}{help ivreg2##swstats:F}({res}{col 17}  1{txt},{res}   318{txt})  P-val{col 37}|{col 39}{help ivreg2##swstats:SW Chi-sq}({res}  1{txt}) P-val{col 60}|{col 62}{help ivreg2##swstats:SW F}({res}{col 67}  1{txt},{res}   318{txt})
{res}cum_lncount_{col 14}{txt}|{col 18}{res}   29.74{col 28}  0.0000{col 37}{txt}|{col 42}{res}   32.47{col 51}  0.0000{col 60}{txt}|{col 65}{res}   29.74

{txt}NB: first-stage test statistics cluster-robust

Stock-Yogo weak ID F test critical values for single endogenous regressor:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for i.i.d. errors only.

{help ivreg2##idtest:Underidentification test}
Ho: matrix of reduced form coefficients has rank=K1-1 (underidentified)
Ha: matrix has rank=K1 (identified)
{res}Kleibergen-Paap rk LM statistic{txt}{col 42}Chi-sq({res}1{txt})={res}23.22  {col 61}{txt}P-val={res}0.0000

{help ivreg2##widtest:Weak identification test}
{txt}Ho: equation is weakly identified
{res}Cragg-Donald Wald F statistic{col 65}   69.57
Kleibergen-Paap Wald rk F statistic{col 65}   29.74

{txt}Stock-Yogo weak ID test critical values for K1=1 and L1=1:
{res}{txt}{col 36}10% maximal IV size{res}{col 67} 16.38
{txt}{col 36}15% maximal IV size{res}{col 67}  8.96
{txt}{col 36}20% maximal IV size{res}{col 67}  6.66
{txt}{col 36}25% maximal IV size{res}{col 67}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.

{help ivreg2##wirobust:Weak-instrument-robust inference}
Tests of joint significance of endogenous regressors B1 in main equation
Ho: B1=0 and orthogonality conditions are valid
{res}Anderson-Rubin Wald test{txt}{col 36}F({res}1{txt},{res}318{txt})={col 49}{res}   0.63{col 61}{txt}P-val={res}0.4294
Anderson-Rubin Wald test{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.68{col 61}{txt}P-val={res}0.4084
Stock-Wright LM S statistic{txt}{col 36}Chi-sq({res}1{txt})={col 49}{res}   0.62{col 61}{txt}P-val={res}0.4296

{txt}NB: Underidentification, weak identification and weak-identification-robust
    test statistics cluster-robust

Number of clusters             N_clust  = {res}       319
{txt}Number of observations               N  = {res}      1270
{txt}Number of regressors                 K  = {res}         1
{txt}Number of endogenous regressors      K1 = {res}         1
{txt}Number of instruments                L  = {res}         1
{txt}Number of excluded instruments       L1 = {res}         1

{txt}HDFE IV (2SLS) estimation
{hline 25}

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on district_fixed

Number of clusters (district_fixed) = {col 33}{res}   319{txt}{col 55}Number of obs = {res}    1270
{txt}{col 55}F(  1,   318) = {res}    0.63
{txt}{col 55}Prob > F      = {res}  0.4293
{txt}Total (centered) SS     = {res} 5.371855681{txt}{col 55}Centered R2   = {res}  0.9640
{txt}Total (uncentered) SS   = {res} 5.371855681{txt}{col 55}Uncentered R2 = {res}       .
{txt}Residual SS             = {res} 5.396645528{txt}{col 55}Root MSE      = {res}  .07982

{txt}{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 21}{c |}{col 33}    Robust
{col 1}          lnturnout{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
cum_lncount_turbine {c |}{col 21}{res}{space 2} -.010717{col 33}{space 2} .0135411{col 44}{space 1}   -0.79{col 53}{space 3}0.429{col 61}{space 4}-.0373585{col 74}{space 3} .0159245
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{help ivreg2##idtest:Underidentification test} (Kleibergen-Paap rk LM statistic):{res}{col 71}  23.216
{txt}{col 52}Chi-sq({res}1{txt}) P-val =  {res}{col 73}0.0000
{txt}{hline 78}
{help ivreg2##widtest:Weak identification test} (Cragg-Donald Wald F statistic):{res}{col 71}  69.568
{txt}                         (Kleibergen-Paap rk Wald F statistic):{res}{col 71}  29.742
{txt}Stock-Yogo weak ID test critical values:{res}{txt}{col 42}10% maximal IV size{res}{col 73} 16.38
{txt}{col 42}15% maximal IV size{res}{col 73}  8.96
{txt}{col 42}20% maximal IV size{res}{col 73}  6.66
{txt}{col 42}25% maximal IV size{res}{col 73}  5.53
{txt}Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
{hline 78}
{help ivreg2##overidtests:Hansen J statistic} (overidentification test of all instruments):{res}{col 71}   0.000
{txt}{col 50}(equation exactly identified)
{hline 78}
Instrumented:{col 23}cum_lncount_turbine
Excluded instruments:{col 23}inter
{hline 78}

Absorbed degrees of freedom:
{hline 20}{c TT}{hline 49}{c TRC}
        Absorbed FE {c |}  Num. Coefs.  =   Categories  -   Redundant     {c |} 
{hline 20}{c +}{hline 49}{c RT}
    stateyear_fixed {c |}          103             103              0     {c |} 
     district_fixed {c |}            0             319            319 *   {c |} 
{hline 20}{c BT}{hline 49}{c BRC}
* = fixed effect nested within cluster; treated as redundant for DoF computation
{res}{txt}(results {stata estimates replay reghdfe_first1:reghdfe_first1} are active now)
(results {stata estimates replay reghdfe_ols:reghdfe_ols} are active now)
(results {stata estimates replay reghdfe_reduced:reghdfe_reduced} are active now)

{com}. 
. 
. *--------------------------- Export LaTeX tables -----------------------------*
. cd "$rootDir/$resultDir/Tables"
{res}/Users/AliceZhang/Dropbox/Research_Columbia/Renewables Voting (Urpelainen Zhang)/JOP/UZ_JOP2021_Replication/Results/Tables
{txt}
{com}. 
. ** Compare OLS and IV estimates
. * OLS
. esttab dem_capacity_ols dem_count_ols rep_capacity_ols rep_count_ols ///
>                 inc_capacity_ols inc_count_ols using TableA11.tex, booktabs replace ///
>                 refcat(cum_capacity_turbine "\emph{c -(}Panel A: OLS{c )-}", nolabel) ///
>                 b(%9.3f) se noconstant noobs nonotes star(* 0.10 ** 0.05 *** 0.01) ///
>                 varlabels(cum_capacity_turbine "Cumulative capacity (MW)" cum_count_turbine "Cumulative count") varwidth(27) modelwidth(13) ///
>                 mtitles("Model" "Model" "Model" "Model" "Model" "Model") ///
>                 mgroups("Democratic Vote" "Republican Vote" "Incumbent Vote", pattern(1 0 1 0 1 0) prefix(\multicolumn{c -(}@span{c )-}{c -(}c{c )-}{c -(}) suffix({c )-}) span erepeat(\cmidrule(lr){c -(}@span{c )-})) ///
>                 width(\hsize)
{res}{txt}(output written to {browse  `"TableA11.tex"'})

{com}.                 
. * IV
. esttab dem_capacity_iv dem_count_iv rep_capacity_iv rep_count_iv ///
>                 inc_capacity_iv inc_count_iv using TableA11.tex, booktabs append ///
>                 nomtitles se noconstant nonotes legend nonumbers collabels(none) star(* 0.10 ** 0.05 *** 0.01) ///
>                 b(%9.3f) stats(N N_clust r2, labels("Observations" "Districts" "\(R^{c -(}2{c )-}\)") fmt(0 0 2)) ///
>                 varlabels(cum_capacity_turbine "Cumulative capacity (MW)" cum_count_turbine "Cumulative count") varwidth(27) modelwidth(13) ///
>                 refcat(cum_capacity_turbine "\emph{c -(}Panel B: IV{c )-}", nolabel) ///
>                 width(\hsize)
{res}{txt}(output written to {browse  `"TableA11.tex"'})

{com}. 
.                 
. ** Close log file
. log close
      {txt}name:  {res}<unnamed>
       {txt}log:  {res}/Users/AliceZhang/Dropbox/Research_Columbia/Renewables Voting (Urpelainen Zhang)/JOP/UZ_JOP2021_Replication/Analysis/logSTATA/012_analysis_Texas.smcl
  {txt}log type:  {res}smcl
 {txt}closed on:  {res} 6 Nov 2021, 20:34:36
{txt}{.-}
{smcl}
{txt}{sf}{ul off}